Orforglipron 36 mg outperforms 12 mg on weight and glycaemia with similar tolerability

Design

Systematic review and fixed-effects meta-analysis of six RCTs

Population

3,459 adults with obesity, with or without type 2 diabetes

Primary outcome

Percentage body weight change, 36 mg vs 12 mg

Effect

Greater reduction with 36 mg (p < 0.00001); adverse events comparable

This meta-analysis pooled six randomised trials (3,459 adults with obesity, with or without type 2 diabetes) that compared the two most studied maintenance doses of orforglipron, the oral non-peptide GLP-1 receptor agonist. It used a fixed-effects model with subgroups by diabetes status and treatment duration.

The 36 mg dose gave greater reductions than 12 mg in percentage and absolute body weight, BMI, HbA1c, waist circumference, triglycerides and systolic blood pressure. Overall adverse events, gastrointestinal events and pancreatic measures did not differ significantly. Small rises in ALT and AST were seen with 36 mg, and pulse rate differed slightly.

The abstract reports p-values rather than the size of the differences, so how much extra weight loss the higher dose buys cannot be stated from it. Orforglipron's regulatory status in India is not established here; treat this as dose-response information for when it becomes available.

• If orforglipron is used, expect more weight and HbA1c reduction at 36 mg than 12 mg.
• Gastrointestinal adverse events were not significantly worse at 36 mg in pooled trials.
• Check liver enzymes when escalating — small ALT and AST rises were seen at 36 mg.
• The pooled trials were short and not powered for rare harms or cardiovascular outcomes.
• Orforglipron is an oral tablet with no food or water restrictions, unlike oral semaglutide.

Why it matters

It suggests the higher maintenance dose is the one to aim for, as with injectable GLP-1 receptor agonists.

Don't overread it

The abstract gives no effect sizes, and fixed-effects pooling of six trials can overstate precision.

The statistics, in plain English

Very small p-values (below 0.00001) say the difference is unlikely to be chance, but not how large it is — a tiny difference in a large pooled sample can be highly significant. A fixed-effects model assumes all trials estimate one true effect; if the trials differ, this makes the confidence look tighter than it should.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.